Nine Dimensions and Zero: Data Discipline Amid the Transfer-Window Noise

On a Tuesday night in Boston, the summer transfer window was at its peak. A long document arrived in my inbox, its subject line reading: "Stage-2 Deep Professional Analysis — Esports Domain". The sender was a data platform I had worked with a few years earlier. I made coffee, pulled up a chair, and began reading with the mindset of someone about to interrogate data.

Fourteen pages. Nine analytical dimensions. Nine tables. Nine sets of conclusions. Each dimension had a bolded heading, a tidy layout, a "Risk Assessment" section, a "Hidden Information" subsection. At a glance, it looked exactly like the kind of professional report any sporting director would hope to receive before a transfer window.

But as I read closely, field by field, I noticed something unusual. The "Game Title" field read: insufficient information. "Version/Patch": insufficient information. "Teams Involved": insufficient information. "Region": insufficient information. "Sponsorship Revenue": insufficient information. "Key Data Points": insufficient information. Nine dimensions, nine tables, and across all those empty cells a single phrase repeated itself — "N/A – insufficient information".

A flawless analytical framework. And a zero at its center.

That was the moment I understood something about my own profession. Over eighteen years observing the sports and esports industries, I have learned that the most dangerous thing is not a lack of data. The most dangerous thing is a framework built beautifully enough to make people forget that there is nothing inside it.

To understand why that document kept me awake, I need to say a little about how sports analytics actually runs. Any serious analytical process moves through two stages. Stage one handles extraction: it reads the source article, pulls out the information points, identifies the core viewpoint, lists the entities mentioned — teams, players, tournaments, games. Stage two is where deep analysis happens: it takes the raw material from stage one, lays it on the operating table, compares, cross-references, builds risk models, sketches scenarios.

The logic is simple: you cannot dissect what has not been placed on the table. If stage one returns a blank page — no title, no source, no information points, no entities — then stage two has nothing to analyze. It can do exactly one thing: admit that it can do nothing.

The document on my desk did precisely that. It moved through nine dimensions, and in each one, instead of inventing an answer, it stated plainly: insufficient information. That sounds trivial. But in a moment when anyone with a language model can produce fluent "analysis" about anything, daring to write "I don't know" becomes a rare act of discipline.

I live in Boston and work as a data consultant for football clubs, but my cradle was esports. In 2026 I started out as a competitor and tournament organiser, then moved into esports media. It was from that cradle that I carried a professional habit into football: every conclusion must have a log. Esports records every millisecond — every keystroke, every second of movement, every decision in a team fight. Football is still in the age of the clay tablet, where people record the scoreline and then delude themselves that this is enough.

And that is precisely why I am grateful for that nine-dimension framework. Not because it gave me answers, but because it taught me how to ask questions.

The first dimension is the patch and game system. This is the master dimension, the precondition without which every other dimension collapses. To analyze a patch, the very first task is to identify which game. League of Legends, Dota 2, CS2, Valorant, Honor of Kings — each has its own frame of reference, its own way of reading a patch. A patch in League of Legends is a story about champion power and lane tempo; in CS2 it is about buy economy and map chokepoints; in Dota 2 it is about tower timings and itemisation rhythm. Without a game title, you cannot choose the analytical lens. That is why this dimension comes first, and why, when it is empty, everything behind it is empty too.

I remember the summer of 2026, when I was still a reporting intern in Foxborough. New England Revolution hosted Toronto FC. Toronto held 72% of possession, fired 21 shots, finished with an xG of 2.3 — and lost 0-1 to a single Diego Fagundez goal. My editor asked me to write about "divine inspiration". I pushed back, pulled the data from StatsBomb, and wrote the piece "Toronto deserved to win 3-0 — the result is a lie". It reached 50,000 reads in 24 hours, and the editor had to publish a correction.

The lesson from that night was not in the number. It was this: if I had not had data to cross-check, I would have written about "divine inspiration" — a beautiful, empty, and false word. Data did not just give me a number; it gave me the right to reject a compelling story that had no basis. The result is a lie that time has memorised; xG is the testimony.

A year later came the 2026 World Cup. I was invited to write data for a new sports platform. Before the quarter-finals, I built a PPDA table for all 32 teams. Croatia posted 8.9 — meaning they allowed opponents an average of only 8.9 passes per defensive action, the lowest of the eight remaining teams. I wrote about Marcelo Brozović: 13.8 kilometres run, nine ball recoveries against Argentina. I posed the question: "Croatia does not have luck, Croatia has a system".

When they reached the final, people started naming me. A Championship club invited me to work as a part-time data consultant. But what I carried away was not a correct prediction. What I carried away was a method: every judgement must come with a threshold. I was no longer afraid to use formulas to strip out the "wrong feeling" of the football public. Croatia's 2026 PPDA board did not measure pressure; it measured pride.

In 2026, the pandemic froze the world, and the stadiums stood empty. The Boston consulting firm where I worked as a mid-level employee cut 40% of its staff. I did not ask to be spared. I wrote a report titled "The Stand Effect: Evidence from 372 Bundesliga Matches Before and During COVID". The numbers showed home-win rates falling from 45% to 31% and penalties dropping by 28%. Huddersfield Town hired me to consult for the final eight Championship rounds. I proposed a rotation model based on sprint distance above 6m/s; anyone running below 80% of the threshold for two consecutive matches had to sit out. They took 14 of 24 points and survived relegation by exactly one point.

The empty stadiums of 2026 were a natural experiment: football does not need an audience to expose its nature. But what mattered more is this: without 372 matches of data to compare before and after, I would simply have been someone sitting around lamenting the pandemic. A natural experiment is only valuable when you have both a control group and a treatment group.

Then came the 2026 Qatar World Cup. I published a series of pre-tournament pieces under the title "Morocco does not defend, Morocco runs data". I pointed out that goalkeeper Yassine Bounou had saved shots worth +4.3 more xG than expected, and that Achraf Hakimi completed 6.8 progressive passes per match. I predicted Morocco would reach the semi-finals. When they beat Portugal 1-0, international platforms called me.

In the 2026 summer window, a Saudi investment fund asked me to appraise Cristiano Ronaldo for a contract extension. I wrote a 40-page report: Ronaldo's true xG output was 0.55, inflated to 0.82 by set-piece situations. I recommended not spending more. The fund objected. Three months later, Ronaldo's market valuation had dropped 15%.

I recount these four stories not to flatter myself. I recount them to make one point: every time I reached a correct conclusion, there was a specific dataset behind it. No dataset, no conclusion. That is the whole content of data discipline.

Back to the nine-dimension document on my desk. The second dimension is tournament system and format. To assess upset probability, you must know what tier the event occupies — Worlds, The International, a Major, a Masters, a regional league, or a tier-two event. You must know the format: single elimination, double elimination, Swiss, groups plus knockouts, or a points system. Each format has a different probability distribution. Single elimination pushes upset probability up; a long points format rewards stability. Without a tournament name and a format, any judgement about "whether the favourites will fall" is mere guesswork.

The third dimension is team and player. This is the heart of all sports analysis. But it demands very concrete things: who was signed, who was released, who went out on loan, who was promoted from the academy, who retired. It demands player names, positions, form curves, age sensitivity. It demands information about the coaching staff and the performance team. When all of that is empty, the "Roster Assessment" table becomes nothing more than a meaningless grid.

The fourth dimension is the regional landscape. The fifth is club finance and business. The sixth is rules and governance. The seventh is the risk profile. The eighth is public narrative and expectation. The ninth is the transmission of the esports industry. Nine dimensions, each a lens, and each lens needs a concrete entity to look at.

What is striking is that the document never pretended. It did not invent a team name, imagine a patch, or attach any transfer deal. In the finance dimension, it stated plainly: "Revenue structure cannot be decomposed because no financial event is described". In the risk dimension, it stated: "No risk can be rated because no subject exists to attach a risk to". In its comprehensive assessment, it delivered a blunt verdict: this is a data-integrity failure at stage one, not an esports event with low information density.

That distinction matters more than it appears. An event with low information density is still an event — there is still something to say, just less of it. An empty input has nothing to say at all. Confusing the two is the origin of everything worthless in this industry.

The document also offered three risk warnings, ordered by priority. First, at a high level: input-integrity failure — stage one returned an empty extraction. Second, also high: downstream contamination risk — any analyst asked to "analyze" an empty input may hallucinate entities or patch details. Third, at a medium level: domain mislabelling risk — the "esports" label was assigned but no esports marker existed.

The second warning is the one that kept me thinking. It speaks to a very human temptation: when handed an empty framework, the natural reflex is to fill it. Human beings are uncomfortable with white space. Analysts are no different. And with a language model in hand, filling white space becomes dangerously easy — it produces sentences that are fluent, confident, and entirely unfounded.

This is where I have to speak about the dark side of the data-analytics profession. We tend to praise the ability to "read a lot from little data". But the line between "reading a lot" and "making it up" is razor-thin. A good analyst is someone who knows exactly how much data is required to reach a conclusion, and who knows how to refuse a conclusion before that threshold is met.

I call it the minimum-viability threshold. Before allowing any deep analysis to run, you need at least one game title, one entity, and one data point. Those three things are the condition for survival. Without them, everything downstream is decoration.

In football, my minimum-viability threshold is even stricter. I never issue a conclusion about a team's form based on fewer than five matches, unless there is a clear structural event — a managerial change, a lost key player through injury, a switch of system. Below five matches, every number sits within the noise band. And noise is not signal.

xG does not judge anyone; it merely exposes the truth that the result conceals. But xG has its own threshold too. A single match with an xG of 2.3 but a 0-1 defeat is a story about luck. Ten matches with high xG yet repeated defeats is a story about a finishing system. Distinguishing those two stories is the entire difference between an analyst and someone who watches highlights.

I have never quit my data addiction; I have only switched suppliers. From being addicted to scorelines, I switched to being addicted to xG. From being addicted to league tables, I switched to being addicted to expected metrics. And at some point, I realised I needed to be addicted to emptiness as well — addicted to the moment when I am allowed to say "I don't have enough data yet".

The transfer window is the most toxic environment for that discipline. Hundreds of rumours appear every day. Each rumour has an unnamed source. Each unnamed source has a motive. An agent leaks to inflate a price. A club leaks to apply pressure. A journalist leaks to win clicks. Amid that chaos, noise drowns out signal to the point where fans lose the ability to tell information from bait.

Transfer data is like a tide: you cannot understand it by looking at the surface; you must measure the seabed. The surface is tweets, sensational headlines, "sources close to the situation". The seabed is contract structure, release clauses, wage bills, age, injury history. Fans look at the surface. Analysts must dive to the seabed.

And that is why the nine-dimension document has value. It refused to look at the surface. It invented no rumour to fill the white space. It dove down, found the seabed empty, and surfaced to report exactly that.

Now comes the contrarian part, the part I believe matters most.

The first thing to say is this: correlation is not causation. This is the cliché of every statistics course, yet it has never stopped being true. A team winning many matches after changing its kit does not mean the kit brings victory. A player scoring many goals after changing his boots does not mean the boots help him score. Sports is riddled with false correlations presented as causation, because causation sells better than correlation.

But there is a deeper contrarian layer that the document revealed. It is this: the more complete a framework is, the more easily it manufactures a false sense of security. Nine dimensions sound very professional. Nine tables sound very thorough. Nine sets of conclusions sound very complete. And precisely because the structure is so beautiful, a reader skimming it might believe they are holding a trustworthy analysis — when in fact they are holding a confession that there is nothing to analyse.

This is the paradox of the template. A template does not lie. A template merely makes people forget to ask questions. When a report has enough headings, enough sections, enough tables, people assume the content inside is equally full. But a beautiful box does not guarantee goods inside.

I have witnessed this paradox in esports many times. A team publishes a player-scouting sheet with full metrics: win rate, KDA, damage per minute. Looking at it, anyone would think they have done their homework. But on closer inspection, those metrics were drawn from too small a sample, or from an outdated game version, or from matches against weak opponents. Win rate does not reveal true skill when the sample is skewed. This is the lesson I carried from esports into football: a correct number in the wrong context is still a wrong number.

The second contrarian layer concerns domain mislabelling. The document carried an "esports" label but contained no esports marker whatsoever — no game title, no team, no player, no tournament. This raises a larger question about how our industry classifies information. When a label is assigned by default rather than by verification, it becomes an accidental lie. And accidental lies are often more dangerous than deliberate ones, because no one checks them.

I think of how sports platforms tag news. A story about a football player might be tagged "esports" merely because the player once played an online game. A story about a basketball league might be tagged "football" because the organiser shares a name. These small errors accumulate into a distorted data system, in which readers can no longer trust the label itself.

The third contrarian layer is about the value of a negative control. In science, a negative-control experiment — one that yields an empty result — has value equal to a positive one. It proves the process works: when there is no agent, there is no reaction. In sports analytics, an empty report plays exactly that role. It proves that the system knows how to refuse when raw material is missing, instead of inventing results.

If you run an analytics department, treasure your empty reports. They are proof that your discipline remains intact. A department that has never returned an empty result is a department making things up.

So what is the signal for the next cycle?

The document left three signals to track. First, the result of re-extracting at stage one — if the information-points field becomes non-empty, the full deep analysis will be triggered. Second, the availability of the source article — if the original text can be recovered, we will know whether the domain label is valid. Third, verification of the domain label — cross-checking the source against esports markers; if a game title, a team, or a player appears, the "esports" classification will be confirmed or corrected.

These three signals sound technical, but they reflect a universal principle: never trust a conclusion without tracing it back to its raw material. Every conclusion must have a path back to the dataset that produced it. When that path is cut, the conclusion becomes a belief, and belief does not belong to analysis.

In the transfer window, this principle matters even more. Every day, a new deal is announced. Every deal carries a transfer-fee figure. And every figure needs verification. Where did that figure come from? From a release clause in a contract, or from an anonymous tweet? From an audited financial statement, or from a rumour passed by word of mouth? The difference between those two kinds of sources is the entire difference between data and rumour.

PPDA in 2026 taught me: pressing is not about running a lot, but running at the right moment. And the transfer window taught me something similar: following the news is not about reading a lot, but reading the right place. Reading the right place means reading contract structure, not headlines. Reading the right place means reading the wage bill, not the fee. Reading the right place means reading injury history, not the trophy cabinet.

I want to close with a progressive thought, not a summary.

Sports is entering a period in which the capacity to produce content has far outstripped the capacity to verify it. Anyone can generate fluent analysis of any match, in any sport, at any time. That abundance is a miracle and a threat at once. It is a miracle because knowledge has become more accessible than ever. It is a threat because the line between analysis and performance grows fainter by the day.

In that world, the greatest value of an analyst is not the ability to write a lot. It is the ability to refuse to write when there is nothing to write. The discipline of saying "I don't know" will become the scarcest asset. Because in a sea of fluent information, the only thing that proves honesty is a white space left intact.

So instead of asking "what does this analysis say", let us learn to ask a different question: "what data does this analysis rest on, and does that data actually exist". That question will filter out most of what circulates online. It will make you skip many compelling headlines. But it will also keep you from believing in numbers born out of thin air.

The nine-dimension document on my desk gave me no conclusion about any team, player, or tournament. It gave me something else, perhaps more valuable: a mirror. In it, I saw reflected the entire temptation of this profession — the temptation to fill white space with beautiful stories. And I saw the reason I must keep refusing that temptation.

The transfer window is long. Thousands more rumours will come. Hundreds more reports will be built with full headings and tables. And I will still be sitting here, in Boston, making coffee, opening each document, and asking one question before reading on: where is the data.

If the answer is that there is none, I will close the document and write a single line: insufficient information. That is not the failure of analysis. That is analysis doing its job.